1. Field of the Invention
This invention relates to an incremental measuring system, such as a system for measuring lengths, in which a measuring scale provided on a scale member is scanned, periodic, basically sinusoidal analog measured-value signals are generated as a result of said scanning, and digital countable signals, which depend on the direction of the scanning movement, are derived from said analog measured-value signals in digitalizing means, preferably by means of trigger and pulse-shaping stages, and are delivered to evaluating means, such as indicating or control means. Electronic means are provided for correcting errors consisting at least of defects of the scale or of the positioning of the scale member in accordance with a stored or predetermined correction program.
2. Description of the Prior Art
In the operation of incremental measuring systems of the kind outlined hereinbefore, a properly designed incremental scale is scanned by optoelectronic, capacitive, inductive or magnetic scanning methods. In most cases two measured-value signals, which are displaced 90.degree. in phase, are delivered to the evaluating means and in most cases a period of the measured-value signals corresponds to an increment of the scale. If incremental scales are used which are optically or opto-electronically readable the measuring scale will preferably consist of bright and dark fields, which are equal in width and are scanned by means of scanning gratings, which have the same increments but are offset from each other by a fractional part of an increment, and by means of illuminating means and photoelectric receivers associated with said gratings. Four photoelectric receivers which are connected in pairs in a back-to-back connection are preferably used to generate two analog measured-value signals.
CH-A-407,569 discloses how the scale can be electronically subdivided by a dividing circuit, which comprises a potentiometer circuit and in which the zero crossings or the slope reversals of the analog measured-value signals appearing at several potentiometer taps are detected by the trigger stages. Dividing circuits are often used which multiply the number of measured-value signals by five so that a slope reversal of the square-wave measured-value signals will be effected at the several trigger stages at intervals of 18, with reference to the analog measured-value signal. The signals resulting from that division are logically combined to produce the digital countable signals, which are delivered to a counter operating a sense which depends on the direction of the scanning movement. The direction of the scanning movement can be detected because one or the other of the analog measured-value signals leads the other signal in dependence on the direction of the scanning movement.
Other means which are known for an electronic subdivision of the scale comprise circuits comprising interpolating computers or combinations of interpolating computers and counters.
Errors may occur in a measurement of lengths and angles and will adversely affect the result of the measurement, particularly in the case of measurements or of a control of machines at high accuracy, unless said errors are corrected. Position errors may often result from errors of the increments of the incremental scale or from errors in the mounting of a system for measuring lengths on an associated machine and/or from errors of the machine. Other errors, which in most cases are linear, may enter the measurement as a result of differential thermal expansion of a scale member which is provided with a longitudinal scale and a machine bed and as a result of a deviation of the scale from an exact parallelism relative to an associated track of a machine.
Known electronic correcting means comprise simple circuits, by which the count that is delivered to the indicating or control means is corrected by a value which is stored in a correction table for each count. A disadvantage of said methods resides in that the count will be suddenly changed for each correction by at least one countable step. In an extreme case, if a correction requires a reversal of the direction of movement and a repeated approach of the scanning unit to the point which is to be measured, a correction by a plurality of countable steps may be required at a correction point. Owing to said sudden changes of the count, a continuous counting is not possible and, as a result, an equivalent value (which has been obtained by a continual counting) will not be obtained for certain positions of the scanning unit. In machine controls which are provided with such correcting means that fact might result in severe disturbances in the operation. It is known that these disadvantages can be eliminated in that digital countable signals are suppressed or additional ones are introduced before the input of the counting stage, e.g., in the pulse-shaping stage. But even in that case there will be sudden changes at the correction points and will result in measurement errors at least of an order of that fractional part of a scale increment which corresponds to one or more countable steps. A correction in dependence on stored correcting values and with the aid of an interpolating computer will require the system for measuring lengths to include a suitable computer and nevertheless may result in inaccuracies.
As has been mentioned hereinbefore, most electronic correcting methods used in practice involve inaccuracies of an order of a length or angle which is associated with at least one countable step and correspondingly large changes of the count must often be tolerated. For this reason, mechanical correcting means have previously been preferred where measurements and corrections of high accuracy were required. Those means are based on the principle that the scanning unit or the scanning grating of the scanning unit is adjusted by correcting drives in the direction of the scanning movement relative to the coupling member provided for displacing the scanning unit. The correcting drives scan correcting templates, which are accommodated in a housing that contains the scale member and are individually adjustable and in response to the scanning of said templates the correcting drives advance or retract the scanning unit or the scanning grating relative to that coupling member. Such correcting methods and correcting means are known from DE-C-866,402 and from U.S. Pat. Nos. 3,039,032; 3,182,385; 2,336,550; and 2,985,826. In accordance with German Patent Specification 2,724,858 the scannable correcting template is constituted by a link chain, which is mounted to extend along the scale member and which at its articulated joints is adjustable by eccentrics, which are operable from the outside. In the mechanical correcting methods which have been described, an infinite correction can be effected in each portion of the scale whereas the known electronic correcting means can effect only a stepwise correction. The corrections which may be effected by the mechanical correcting methods are limited by the extent to which the scanning unit or the grating can be adjusted relative to the coupling member, particularly in the case of substantial linear errors, and by the spacing of the points at which the correcting template can be adjusted. The mechanical correction involves a very high expenditure and requires the use of a much larger housing for the measuring system and of a relatively large scanning unit, which must be provided with the required correcting drives and must have a relatively heavy weight, and the mechanical correction cannot be effected unless the measuring system has inherently been designed for a performance of such mechanical corrections. In case of linear scale errors of metal scale members a continuous correction can be effected in that the scale member is elongated or upset. But that practice also involves a considerable expenditure and a severe restriction regarding the design of the measuring system.